In this video, Microsoft Developer’s Ayan Gupta and Julien Dubois guide Java developers through building intelligent AI-powered applications using LangChain4j, Spring Boot, GitHub Copilot, and Azure AI Foundry.

Building Intelligent AI Applications with Java, Spring Boot, and LangChain4j

Presenters: Ayan Gupta (Microsoft) and Julien Dubois (Microsoft Java Developer Advocacy, LangChain4j core contributor)

Introduction

Modern applications increasingly require adaptive intelligence, similar to how a smart mug reacts to temperature changes. This video introduces Java developers to the process of building AI-powered apps using LangChain4j, Spring Boot, GitHub Copilot, and Azure services.

Key Learning Points

  • LangChain4j Overview: Learn why LangChain4j is a preferred framework for Java-based AI integration, offering a unified interface to various LLM providers.
  • Getting Started: Project setup via start.spring.io, configuration with Maven (Java 24), and running a basic Spring Boot application.
  • Feature Addition: Use GitHub Copilot to quickly add a CommandLineRunner for command-line interactivity.
  • Manual User Input Testing: Understand how to test non-AI functionality before integration.

Integrating AI Capabilities

  • Dependency Management: Utilize Maven’s Bill of Materials (BOM) for cleaner management of LangChain4j dependencies across modules.
  • OpenAI SDK Integration: Choose and add the official OpenAI Java SDK as an LLM backend with LangChain4j.
  • Configuring Chat Models: Connect your application to Azure AI Foundry and GPT-4o Mini, specifying the base URL, API key, and model name for the chat model.
  • Unified LLM Interface: Demonstrated usage of the ChatModel interface, which can be pointed at various LLM implementations (OpenAI, Mistral, Llama, or GitHub Models) without code rewrites.

Practical Demonstration

  • Step-by-Step Demo:
    • Add required dependencies with Maven.
    • Configure chat model for Azure AI Foundry integration.
    • Retrieve and set up API keys and endpoints.
    • Craft and test an AI-powered poem generator in Spring Boot.
    • Recap session takeaways and next steps for further exploration.

Resource

Timeline Highlights

  • 0:00 – Introduction to intelligent applications
  • 2:32 – Maven setup (Java 24)
  • 2:52 – Adding features with GitHub Copilot
  • 4:28 – Maven BOM for LangChain4j dependencies
  • 6:27 – Connecting to Azure AI Foundry
  • 8:47 – Live AI prompt creation
  • 10:30 – Recap and future roadmap

Summary

Participants learn to architect and develop AI-driven apps with Java, leveraging Microsoft’s Azure AI services, LangChain4j’s flexible framework, and Copilot-driven productivity to streamline modern development workflows.